# How to get your first users when nobody knows you exist > Getting the first users for a side project when you have no audience: where they actually come from, what order to try, and what to stop doing. Canonical: https://outclimbed.com/blog/first-users-for-a-side-project Growth Outclimbed November 10, 2026 ยท Updated September 13, 2026 Getting the first users for a side project is mostly unscalable work that nobody enjoys writing about. The tactics below are not secret. What separates the people who get first users from the people who keep reading about it is doing the individual half before the broadcast half. ## The order that works 1. Talk to people individually where the problem is already discussed. 2. Make the product findable so the slow channel starts early. 3. Broadcast, once there is somebody to broadcast to. Almost everybody does this backwards, starting with a launch. A launch with no audience is a quiet day, and a quiet day feels like a verdict on the product when it was a verdict on the distribution. ## Go where the problem already gets discussed Not where your buyers hang out in general. Where the specific problem your product solves is being complained about: a subreddit thread, a forum question, a recurring gripe in a Slack group, an issue on a competitor's tracker. Answer the question properly first, without mentioning your product. Mention it only if it is genuinely the answer. This converts at a rate no broadcast channel comes close to, and it does not scale, which is exactly why it is available to you. ## Talk to twenty people individually Twenty messages, written one at a time, referencing something that person actually said or built. Ask one question. Do not pitch in the first message. Most will not reply. Three or four will, and those conversations will tell you more about your positioning than any amount of analytics. Your first users almost always come out of this pile. The test for whether a message is worth sending: could it have been sent to anybody else unchanged? If yes, it is a broadcast wearing a direct message's clothes and it will be read as one. ## Ten who come back beat a thousand who signed up At this stage the number to watch is not signups. It is how many people used the thing more than once without being reminded. Ten of those is a product. A thousand signups with none of them is a landing page that converts. ## Set up the slow channel early Directory listings and search traffic take months to do anything, which is precisely the argument for starting them in month one rather than month six. It is a weekend of forms that pays out while you are doing something else. [Where to submit a SaaS](https://outclimbed.com/blog/where-to-submit-your-saas) sets out an order that tests before it spends, and [what directory submission involves](https://outclimbed.com/blog/startup-directory-submission) covers what the forms actually ask for. ## Where launching fits After the first two, not before. A launch amplifies an audience; it does not create one. If the first twenty conversations produced people who liked the product, a launch gives them somewhere to say so publicly, and that is when launch day works. If it has already gone flat, [what to do after a quiet launch](https://outclimbed.com/blog/product-hunt-launch-failed) is about salvaging the part that still has value, which is the comments. The same goes for a seasonal sale. A Halloween or Black Friday discount gives people who already know the product a reason to act now; it does not find new ones. If you are thinking of running one, the [Halloween SaaS deals we checked](https://outclimbed.com/blog/halloween-saas-deals-2026) show how other small vendors set dates, codes and limits. ## Where first users actually come from Asking founders after the fact, the answers cluster into four, and the proportions are not what most people expect. **People the founder already knew.** Former colleagues, people from a previous project, somebody who follows them for unrelated reasons. Not a scalable channel and not a fake one: these are real users with a real problem and they count. **A specific thread or conversation.** One answer in one place where the problem was being discussed. This is where the first users of a surprising number of products came from, and it is the one thing on this list that is available to somebody with no audience at all. **Search, months later.** Slow, cumulative, and the one that is still working in year two. This is the argument for setting up listings and content early even though nothing happens for a while. **A launch.** Real, but usually amplifying an audience the founder already had rather than creating one. ## What to actually say to somebody The blocker for most people is not knowing what a good first message looks like. It is shorter than you think. Name the specific thing they said or built, in one sentence, so it is obvious this was not sent to a hundred people. Say what you made in one sentence, plainly, without adjectives. Ask one question that they can answer in a line. Stop. Do not attach anything, do not ask for a call, and do not include a pitch deck. The goal of the first message is a reply, not a customer. The customers come from the third or fourth message in a conversation that started because somebody felt like answering. ## What to stop doing - **Posting into the void.** Broadcasting to an account with no followers is not marketing, it is journalling. - **Rebuilding the landing page again.** If ten people have not used the product, the landing page is not the constraint. - **Waiting for it to be ready.** The first users are the ones who tell you what ready means. ## The unscalable part is the point The best-known argument for this is [Paul Graham's essay on doing things that do not scale](https://paulgraham.com/ds.html), and it is worth reading in full because the usual summary loses the mechanism. Recruiting your first users by hand is not a stopgap you tolerate until real marketing arrives. It is how you find out what the product should be. An automated channel gives you numbers. Twenty conversations give you the sentences people use for the problem, which is what your landing page should say, which is what makes the automated channel work later. Skipping to the automated channel means guessing at those sentences, and a channel pointed at the wrong words converts badly no matter how much is spent on it. There is also a compounding effect nobody mentions. The people who become your first users this way remember that a person contacted them individually. They reply to emails later, they answer questions, and a disproportionate number of them tell somebody else. A cohort of first users acquired by hand behaves differently for a year afterwards. ## How long this should take Longer than a launch and shorter than people fear. Twenty individual messages is an afternoon. Getting ten people who come back is usually several weeks of iterating on what those conversations told you, not months. What takes months is the slow channel, which is exactly why it starts in parallel rather than afterwards. If the directory listings and the first content go up in week one, they are producing something by the time the hand-recruited first users have told you what the product actually is. ## What to measure in the first month Three numbers, and none of them is traffic: how many people used it twice, how many conversations you had with a human, and where each returning user came from. The third one requires tagging your links, which takes a minute and is the difference between knowing and guessing. ## Where a directory fits into this outclimbed, the site this is on, is a paid directory, so treat this as disclosed rather than neutral. It belongs in step two: a place starts at $10 and it is a background channel rather than a launch. Position can also be earned by sending people rather than by paying more, which is the part built for someone who has no budget at this stage. It will not get you your first ten users. The twenty messages will. [About](https://outclimbed.com/about) is straightforward about what a paid place does and does not buy. ## The trap to watch for The most expensive failure mode here is substituting preparation for contact. Rebuilding the landing page, redesigning the logo, adding the feature you are sure is blocking adoption: all of it feels like progress and none of it involves a person who might use the thing. The test is simple and slightly uncomfortable. At the end of a working week, how many individual humans did you speak to about the product? If the answer is zero for three weeks running, you are not looking for first users, you are avoiding the part where somebody says no. One more thing worth saying about the emotional side, because it decides whether people keep going. Hearing no from twenty individual people feels far worse than getting no signups from a thousand anonymous visitors, even though the second one is a worse result. The anonymous version lets you believe the problem was the headline. The individual version tells you exactly what was wrong, which stings and is the entire value. The founders who find their first users are not the ones who guessed better. They are the ones who were willing to hear the specific version of no often enough to learn from it, and who kept the slow channels running in the background while they did. ## Questions **How many first users do I need?** Ten who come back. **Should I launch first?** Only if you already have people to bring. **Do directories help?** Slowly, which is why they should be set up early rather than relied on. outclimbed is a directory where every position has a public price, and where a share link moves you up without spending again. [Start a climb - $10](https://outclimbed.com/enter) [See the wall](https://outclimbed.com/) ## Common questions How do I get the first users for a side project? By going individually to places where the problem is already being discussed and talking to people, then making the product findable through directories and search. The individual half comes first and is the part most people skip. How many first users do I need? Ten who use it repeatedly tell you more than a thousand who signed up once. The goal at this stage is information, not a number. Should I launch to get my first users? A launch works when you already have an audience to mobilise. Without one it tends to produce a quiet day, which is why the unscalable outreach usually has to come first. Do directories help get early users? They help steadily rather than immediately. Directory traffic arrives from search over months, which makes it a poor launch tactic and a good background one to set up early. ## Read next - [Your Product Hunt launch flopped. What to do next](https://outclimbed.com/blog/product-hunt-launch-failed) A flat Product Hunt launch is not a verdict on your product. What the result actually measures, what to salvage from the day, and where to go next. - [Where to submit your SaaS, and in what order](https://outclimbed.com/blog/where-to-submit-your-saas) Where to submit your SaaS without wasting a week on it: which places are worth the form, what each one wants, and how to tell when to stop submitting. ## Related - [Blog](https://outclimbed.com/blog) - [Your Product Hunt launch flopped. What to do next](https://outclimbed.com/blog/product-hunt-launch-failed) - [Where to submit your SaaS, and in what order](https://outclimbed.com/blog/where-to-submit-your-saas) - [Halloween SaaS deals 2026: every discount checked on the vendor's own page](https://outclimbed.com/blog/halloween-saas-deals-2026) - [The wall: the live ranking](https://outclimbed.com/)